Posted January 24, 2017 Just now, Yldrania said: Then I'm not sure what is going wrong, but if it works for others that only means it should be fixable in some way. They are the biggest server hoster in Europe so they will be able to fix this for sure, but if no one keeps in touch with them, they can't solve the problem. Communication! Share this post Link to post Share on other sites
Posted January 24, 2017 Just now, Sklo:D said: They are the biggest server hoster in Europe so they will be able to fix this for sure, but if no one keeps in touch with them, they can't solve the problem. Communication! That's what I mean. If there is a fix that is possible to get from this server, I believe that the devs should do whatever necessary to actually fix this issue, since it not only makes Wurm unenjoyable but also causes a decrease in player numbers, which the game really can't afford. Share this post Link to post Share on other sites
Posted January 24, 2017 (edited) There is an official guide telling what to do when such things happen, no one from the dev team mentioned it yet. Using ping plotter shows exactly what's wrong! https://wiki.hetzner.de/index.php/Pingplotter/en Edited January 24, 2017 by Sklo:D Share this post Link to post Share on other sites
Posted January 26, 2017 I have been using pingplotter to keep an eye on the router I posted above, it spends about 12-15 hrs between 2k-3k ms. From my workstation. From the vpn connection to paris (images will not insert correctly.) This is the only way I can play wurm online. Share this post Link to post Share on other sites
Posted January 26, 2017 (edited) Ping plotter is just a pretty picture maker. It still uses standard protocols under the hood and a standard tcp traceroute (like my original post) clearly shows that one of their hops has a serious config issue. 4 seconds is an eternity for a router, it's not a buffering/bandwidth issue it's a config issue. So sure if they want pingplotter output that's fine, but any decent network engineer should be able to see a single traceroute report with those numbers and know that they had an issue and would then set up their own test points and begin probing in a systematic way till it's resolved. So sorry, but if Hetzner needs 100s of people to submit ad hoc pingplotter results from god knows where to resolve this then they may be the biggest but they're incompetent. I think the more likely scenario is they don't care. If one person reports they can ignore it, if 10's of people report they can ignore it, if 100s of people report they can still ignore it. If people take their money elsewhere, well that's different. Edited January 26, 2017 by belthize Share this post Link to post Share on other sites
Posted January 28, 2017 (edited) On 1/25/2017 at 5:32 AM, Sklo:D said: Promise you there is no better host than hetzner. Others will just make more problems, also hetzner provides a very good DDOS protection, which is priceless. My server has the same amount of players often more than deli, exo or the whole epic cluster, etc, and we are having almost no issues. Very stable and powerful connection, no noticeable lag most of the time. You are in Austria and hetzner peer at the Vienna Internet Exchange so sure, hetzner is probably fine for you, they deliver the traffic locally to your country them selves, but for people on the other side of the world they are terrible. It's not necessarily hetzner's fault directly it's that their choice of upstream providers (TATA / GTT) mean there is poor service for people far away. If Hetzner peered in north america there would likely be no issue because it would prevent and issues from crossing the atlantic or crossing USA. Other hosts for example OVH, Softlayer, Leaseweb peer at internet exchanges all over the world in North America, Europe, Asia, Australia etc so they control the path very close to the user, many of them also choose higher quality transit providers for example Level3 and NTT. Furthermore in the case of OVH their entire network traffic status is visible 24/7 and you rarely see any congestion http://weathermap.ovh.net/ On 1/25/2017 at 5:35 AM, Sklo:D said: They are the biggest server hoster in Europe Not by a long shot. Hetzner only have 2 DC's in Europe, Leaseweb have 9, OVH have 18 with like 10 more under construction, Softlayer have 20 OVH have over 200,000 servers in Europe alone, Leaseweb something like 140,000 - Hetzner does not even have 100,000. Hetzner's network capacity is around 1.5 tbps. OVH's is over 10 tbps. Softlayer is probably 10+ tbps since IBM bought them. Edited January 28, 2017 by Gavin 4 Share this post Link to post Share on other sites
Posted January 28, 2017 Just to reiterate my routing on my primary ISP has changed, it used to go via HE.net (my ISP's upstream provider of choice, they are fairly good) but obviously hetzner has changed something and they request it travel over as6453 who my ISP peer with in California (as6453 is TATA) and TATA are useless. Tracing route to hetzner.de [213.133.107.227] over a maximum of 30 hops: 1 <1 ms <1 ms <1 ms 192.168.1.1 2 2 ms 2 ms 2 ms 10.20.22.109 3 2 ms 1 ms 2 ms per-pow-stg-csw2-tg-2-4.tpgi.com.au [202.7.180.121] 4 2 ms 2 ms 3 ms per-pow-stg-crt2-port-channel-2.tpgi.com.au [202.7.190.113] 5 54 ms 53 ms 53 ms 202-7-162-125.tpgi.com.au [202.7.162.125] 6 57 ms 59 ms 59 ms 203.29.134-4.tpgi.com.au [203.29.134.4] 7 240 ms 242 ms 233 ms ix-ae-17-0.tcore2.SQN-San-Jose.as6453.net [64.86.21.53] 8 388 ms 387 ms 387 ms if-ae-1-2.tcore1.SQN-San-Jose.as6453.net [63.243.205.1] 9 * * * Request timed out. 10 390 ms 391 ms 394 ms if-ae-11-2.tcore2.L78-London.as6453.net [63.243.128.38] 11 387 ms 386 ms 386 ms if-ae-8-2.tcore2.AV2-Amsterdam.as6453.net [80.231.131.6] 12 384 ms 384 ms 384 ms if-ae-2-2.tcore1.AV2-Amsterdam.as6453.net [195.219.194.5] 13 393 ms 387 ms 393 ms if-ae-6-2.tcore1.FNM-Frankfurt.as6453.net [195.219.156.61] < 390ms to get to central Europe from Australia is pretty good, no problems so far 14 902 ms * * if-ae-7-2.tcore1.FR0-Frankfurt.as6453.net [195.219.50.1] < last hop before it enters Hetzner's network - this could either be a issue on TATA's network in Frankfurt, or it's congestion on hetzner's TATA transit link 15 891 ms 972 ms 812 ms 195.219.219.10 16 607 ms 1536 ms 992 ms core11.hetzner.de [213.239.245.34] 17 1494 ms * 868 ms ex9k2.rz1.hetzner.de [213.239.203.214] 18 1025 ms 977 ms 1772 ms www.hetzner.de [213.133.107.227] Trace complete. This is the problem. People who don't live in western Europe can't just enter hetzner's network via some cheap peering link where both your ISP and Hetzner have mountains of idle capacity even in peak time, we travel over international transit networks which are often running at capacity, in many cases intentionally to try and extort paid peering settlements from major carriers. Even thou the above trace shows 18 hops I guarantee you it's gone through 40+ routers that may not show up in traces. Again, as mentioned in my post on page 1 this is why companies like Netflix or Google setup local mirrors in every country they service because you cannot get reliable service to everyone just plugging your network in a few global transit providers, as people say the internet is a series of tubes and many of those tubes are completely broken and blocked up, the only way to get around this is to setup a PROPER network that has points of presence across the globe so you control the tubes yourselves, AKA almost every professional large scale data-center. 2 Share this post Link to post Share on other sites
Posted January 28, 2017 1 2 ms 2 ms 2 ms 2 7 ms 11 ms 7 ms 3 8 ms 7 ms 7 ms 4 44 ms 44 ms 46 ms 5 35 ms 37 ms 37 ms 7 45 ms 45 ms 42 ms dcp-brdr-04.inet.qwest.net [67.14.28.18] 8 34 ms 36 ms 37 ms ix-ae-16-0.tcore2.AEQ-Ashburn.as6453.net [216.6.87.117] 9 134 ms 136 ms 134 ms if-ae-11-2.thar2.NJY-Newark.as6453.net [216.6.87.242] 10 132 ms 134 ms 134 ms if-ae-1-3.thar1.NJY-Newark.as6453.net [216.6.57.1] 11 139 ms 129 ms 128 ms if-ae-8-2.tcore1.LDN-London.as6453.net [66.198.70.174] 12 123 ms 135 ms 136 ms if-ae-17-2.tcore1.L78-London.as6453.net [80.231.130.129] 13 137 ms 136 ms 139 ms if-ae-3-6.tcore1.PYE-Paris.as6453.net [80.231.130.86] 14 134 ms 135 ms 138 ms if-ae-11-2.tcore1.PVU-Paris.as6453.net [80.231.153.49] 15 127 ms 127 ms 127 ms if-ae-9-2.tcore2.FNM-Frankfurt.as6453.net [195.219.87.9] 16 135 ms 124 ms 130 ms if-ae-4-2.tcore1.FR0-Frankfurt.as6453.net [195.219.87.18] 17 2956 ms 2983 ms 3103 ms 195.219.219.10 18 3370 ms 3585 ms 3647 ms core23.hetzner.de [213.239.229.74] 19 1887 ms 1855 ms 1832 ms ex9k2.rz20.hetzner.de [213.239.229.10] 20 3521 ms 3399 ms 3089 ms xanadu.game.wurmonline.com [148.251.21.209] Share this post Link to post Share on other sites